24 November 2012


From the critique by Mark Rothstein in 25(5) GeneWatch (2012) 15-16 of the US Presidential Commission for the Study of Bioethical Issues Privacy and Progress in Whole Genome Sequencing report (noted here).

Rothstein comments that -
The report places too much emphasis on research and not enough on clinical applications. The research issues are relatively uncontroversial, because there is widespread societal recognition of the desirability of research utilizing WGS, so long as there is informed consent, data security, and other traditional protections. The clinical applications are less easily resolved. Certain uses of WGS are undoubtedly valuable, including for analyzing rare disorders, performing tumor genome sequencing, and determining pharmacogenomically appropriate medications. For other uses, however, such as predictive risk assessment for common, complex disorders in asymptomatic individuals, there is little current clinical utility and therefore WGS is difficult to justify.
Another weakness of the report is that it underestimates the potential harms that individuals may suffer as a result of WGS. Except for observing the burden of knowing about a condition for which there is no effective treatment, the report focuses solely on tangible harms, such as discrimination. Nevertheless, many individuals currently experience various psychological and social harms from traditional genetic testing (e.g., depression, anxiety), and the massive scope of WGS is likely to increase the number and severity of these intangible problems.
The discussion of privacy, the main focus of the report, is incomplete. By emphasizing the risk of unauthorized uses of genomic information the report overlooks the substantial issue of lawful uses of information pursuant to compelled authorizations. This occurs when individuals are required to sign a broad authorization (releasing substantially all of their health records) as a condition of applying for a job, various forms of insurance, government benefits, or other matters. Each year in the U.S. individuals sign at least 25 million compelled authorizations, and the adoption of interoperable, comprehensive, and longitudinal electronic health records greatly increases the scope of disclosure and therefore the privacy risks. Unless these risks are addressed it is impossible to protect genetic privacy -- or more broadly, informational health privacy.
The report also contains an inadequate (and, in places, incorrect) discussion of applicable laws. For example, it states (on pages 66-67) that GINA “does not address the use of or access to genetic data. In other words, GINA is an anti-discrimination law; it does not provide comprehensive privacy protections.” This statement is incorrect. Section 202 of GINA prohibits an employer from requiring or requesting an individual to undergo genetic testing or disclose genetic information as a condition of employment. In theory, an employer cannot discriminate if it does not have genetic information; also, individuals will be more willing to undergo beneficial testing if the results will not be available to employers. The problem is that there is no practical way for custodians of health records to comply with a request to disclose everything except genetic information (which, under GINA, includes family health information). By calling attention to this problem, the Commission could have helped to hasten the development of necessary privacy-enhancing technologies.
Although the report is limited to broadly defined “privacy” issues, it is impossible to study privacy concerns without understanding the context in which WGS will be used. The adoption of WGS technology has the capacity to overturn numerous established practices in clinical genetics, and the report’s failure even to mention these applications is unfortunate. Three examples follow.
First, genetic screening of newborns and children is now limited to conditions for which medical intervention in childhood is necessary and potentially beneficial. The rationale for limited testing in childhood is that for adult-onset disorders that cannot be ameliorated in childhood (e.g., Huntington disease, Alzheimer’s disease), the child should be able to decide upon reaching maturity whether to undergo genetic testing. Routine, population-wide WGS of newborns and children would fundamentally alter this established policy. Such a change could have significant psycho-social implications, including privacy implications, and therefore requires thoughtful analysis.
Second, as the cost difference between a single genetic test and WGS is reduced to nominal levels, there will be pressure to undergo WGS in every situation where only a single test is needed initially. This “might as well” sequencing could be promoted by public or private payers as being more efficient than multiple tests, as well as by clinicians who believe the additional genomic information has clinical value. WGS will generate numerous incidental findings that would necessitate genetic counseling, surveillance, and privacy controls. It also could lead to a significant psychological burden that needs to be considered.
Third, the report does not address direct-to-consumer WGS. In addition to numerous regulatory issues, direct-to-consumer WGS raises such fundamental ethical issues as the conflict between autonomy and paternalism, nonmaleficence, justice, and privacy.